The Autism Detection in Early Childhood Tool: Level 2 autism spectrum disorder screening in a NICU Follow-up program

Objective: Children born preterm are at increased risk for autism spectrum disorder (ASD). However, early diagnosis of ASD is challenging because conventional screening Level 1 tools are less reliable in this population. We sought to determine whether the Autism Detection in Early Childhood (ADEC) and Child Behavior Checklist (CBCL) could accurately identify children at risk for ASD in a NICU Follow-up setting and thus facilitate referral for formal ASD evaluation.

Method: Children aged 18-36 months were recruited from a NICU Follow-up program. All children received presumptive diagnoses based on DSM-5 criteria and were screened for ASD risk with the ADEC and CBCL. Children scoring in the "at risk" range on either tool were referred for a full diagnostic ASD evaluation.

Results: Sixty-nine patients (median birth weight 1140 g; median gestational age 28 weeks) were included with 18 designated "at risk" for ASD. Nine (13 %) scored "at risk" on the ADEC and 12 (17 %) on the CBCL. Thirteen children underwent diagnostic ASD evaluation with 9 receiving a formal diagnosis of ASD. The ADEC demonstrated the best performance (sensitivity 89 %, specificity 98 %). The CBCL was less sensitive (sensitivity 50 %, specificity 90 %). Requiring elevated scores on both the CBCL and ADEC was specific but not sensitive (sensitivity 33 %, specificity 100 %).

Conclusion: The ADEC performed well in identifying children at risk for ASD within this high-risk NICU cohort, adding benefit as an autism-specific screening tool over the CBCL alone.